Skip to content

Latest commit

 

History

History
68 lines (56 loc) · 4.65 KB

README.md

File metadata and controls

68 lines (56 loc) · 4.65 KB
Raw

Mirzadeh_Cal

R scripts and functions for tidying, visualizing, and analyzing indirect calorimetry data from Sable Promethion.

Please note that these scripts are a work-in-progress, and were written with the specific needs of the Mirzadeh Lab in mind. Therefore, they may require customization for your group's specific needs. I have detailed below the assumptions and requirements for this script to function properly, but don't hesitate to email me (or submit an 'Issue') with any questions, concerns, requests - or pull requests if you've solved or improved a problem!

Requirements:

  • Expedata (.exp) data files processed with MI >v2.46 in 1-minute or 3-minute bins - all files must be processed with the SAME MACRO to ensure that the column names and binnings are consistent.
  • .csv containing the TimeSeries sheet from the macro-processed .xml.
  • This code will work best if you implement the following naming convention: <"yyyy-mm-dd_runid_timeseries_m.csv">. If you name your raw .exp files with the yyyy-mm-dd_runID convention, the One-Click-Macro (OCM) can be modified to automatically export the TimeSeries as a .csv with this naming convention applied. Contact Sable for details on implementing this in your OCM.
  • .csv containing any grouping/identifiers/metadata supplied to Sable in the 'Animal Data Entry' form. This file should have the same filename as the Group decoding sheet with same naming convention as TimeSeries sheet (e.g., 2023-06-08_cal021_KEY.csv), saved in same directory. A template is included in this repository. Until such time as Sable enables automatic export of this metadata from the OCMs, users of this code must manually enter their information into the template, and save it with the appropriate filename in the same folder as the TimeSeries.csv file.

Workflow:

1) Cal_Clean

Tidies and computes hourly bins (for time-series plots) and photoperiod-averaged data frames (for summary boxplots). Assumes a 12/12 light/dark cycle with lights off at 18:00. To ensure each day has a full 24-hours (for photo-period averaging and boxplots), this script also trims the data to always start and stop at 18:00.

  • Required Inputs:
# Assumes file-naming convention of: yyyy-mm-dd_runID.csv
filename        <- '2023-06-08_cal021_timeseries_m.csv'
fpath           <- "C:/path/to/filefolder"
  • Optional Inputs:
NCDgcal         <- 3.35 # default conversion of kcal per gram for normal chow 
HFDgcal         <- 5.47 # default conversion of kcal per gram for high-fat diet
segment.run     <- FALSE # for conditional execution of Cal_Segment 
cols2excl       <- c('Still_pct_M','Sleep_pct_M',
                     'XBreak_R','YBreak_R','Mass_g','AllMeters_M') # default columns to exclude
  • Outputs one .Rda file containing the following:
    • df : Cleaned data frame at original sampling interval as performed by the macro.
    • df.hourly : Cleaned data frame re-sampled into hourly bins.
    • df.env : Data frame containing ESA data
    • df.daily : Cleaned data frame re-sampled into daily photoperiod bins.
    • df.avg.total : Tidy data frame containing data averaged across the entire experiment (mean/total per day)
    • df.cum.total : Somewhat redundant with df.avg.total, but frames the experimental averages as mean/total across experiment, instead of per day.
    • key : Data frame containing run metadata for unblinding investigator.

2) Cal_Inspect

Work in progress: Will be built up as a way to interactively view and assess quality of data.

3) (Optional) Cal_Segment

Work in progress: This was a script requested for splitting large recordings into distinct segments.

4) (Optional) Cal_Merge

Work in progress: This script was requested for merging separate recordings into one data frame. Please contact Chelsea before using.

5) Cal_Plot

Takes the Clean.Rda file as input, and automatically generates time-series and photoperiod-averaged boxplots for the requested variables. User may submit a grouping variable (that should be present within the 'KEY' file - see above), and a faceting variable, to further subset their data into panels. Requires the "Cal_Units.csv" file saved within the same folder as the Clean.Rda file to be loaded.

Mirzadeh_Cal_example_ts Mirzadeh_Cal_example_boxplot2

Please acknowledge use of this repo in any publications.

For questions, issues, suggestions, please submit an issue or contact:

Chelsea Faber
Mirzadeh Laboratory at Barrow Neurological Institute
kasper DOT chelsea AT gmail DOT com
02-Aug-2022